Cu2Zn(Sn1-xGex)Se4 Heterointerface and

Subscriber access provided by Access provided by University of Liverpool Library

Surfaces, Interfaces, and Applications 2

1-x

x

4

Band Alignment of the CdS/CuZn(Sn Ge)Se Heterointerface and Electronic Properties at the CuZn(Sn Ge)Se surface: x = 0, 0.2, 0.4 2

1-x

x

4

Takehiko Nagai, Takuya Shimamura, Kohei Tanigawa, Yuya Iwamoto, Hiroya Hamada, Nobuyoshi Ohta, Shinho Kim, Hitoshi Tampo, Hajime Shibata, Koji Matsubara, Shigeru Niki, and Norio Terada ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b19200 • Publication Date (Web): 09 Jan 2019 Downloaded from http://pubs.acs.org on January 11, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Band Alignment of the CdS/Cu2Zn(Sn1-xGex)Se4 Heterointerface and Electronic Properties at the Cu2Zn(Sn1-xGex)Se4 Surface: x = 0, 0.2, 0.4 Takehiko Nagai,†,* Takuya Shimamura,‡ Kohei Tanigawa,‡ Yuya Iwamoto,‡ Hiroya Hamada,‡ Nobuyoshi Ohta,‡ Shinho Kim,† Hitoshi Tampo,† Hajime Shibata,† Koji Matsubara,† Shigeru Niki,§ and Norio Terada‡ †Research

Center for Photovoltaics (RCPV), National Institute of Advanced Industrial Science

and Technology (AIST), Central 2, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan. ‡Graduate

School of Science and Engineering, Kagoshima University, 1-21-40 Korimoto,

Kagoshima 890-0065, Japan. §

Department of Energy and Environment (E&E), National Institute of Advanced Industrial

Science and Technology (AIST), Central 1, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8568, Japan.

KEYWORDS: Band alignment, CZTGS, IPES, UPS, XPS, Kesterite, Solar cell.

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 41

ABSTRACT

The surface electronic properties of the light absorber and band alignment at the p/n heterointerface are key issues for high performance heterojunction solar cells. We investigated the band alignment of the heterointerface between cadmium sulfide (CdS) and Ge incorporated Cu2ZnSnSe4 (CZTGSe), with Ge/(Ge+Sn) ratios (x) between 0 and 0.4, by X-ray photoelectron, ultra-violet, and inversed photoemission spectroscopies (XPS, UPS, and IPES, respectively). In particular, we used interface-induced band bending in order to determine the conduction-band offset (CBO) and valence-band offset (VBO), which were calculated from the core-level shifts of each element in both the CdS overlayer and the CZTGSe bottom layer. Moreover, the surface electronic properties of CZTGSe were also investigated by laser-irradiated XPS. The CBO at the CdS/CZTGSe heterointerface decreased linearly, from +0.36 to +0.20 eV, as x was increased from 0 to 0.4; in contrast, the VBO at the CdS/CZTGSe heterointerface was independent of Ge content. Both UPS and IPES revealed that the Fermi level at the CZTGSe surface is located near the center of the bandgap. The hole concentration at the CZTGSe surface was of the order of 1011 cm-3, which is much smaller than that of the bulk (~1016 cm-3). We discuss the differences in hole deficiencies near the surface and in the bulk on the basis of laser-irradiated XPS, and conclude that hole deficiencies are due to defects distributed near the surface with densities that are lower than in the bulk, and the Fermi level is not pinned at the CZTGSe surface.


2

Page 3 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Solar cells based on non-toxic and Earth-abundant materials are required in order to renewably produce energy on the multi-terawatt scale.1-4 Thin-film solar cells have the great potential to reduce costs, and their large-scale production will realize a low-carbon-emission society. In particular, Cu2ZnSn(SySe1-y)4 (CZTSSe), with the kesterite crystal structure, is a candidate lightabsorbing material5-12 that can replace those used in recently commercialized solar cells based on Cu(In1-zGaz)Se2 (CIGS) and CdTe, because CZTSSe does not include low-abundant elements such as In, Ga, or Te. Moreover, CZTSSe exhibits strong optical absorption,13-16 and its bandgap energy can be finely tuned over the ~1.0 to ~1.5 eV range by adjusting the anionic mixing ratio (e.g., S/(S+Se)).17,18 The efficiency of CZTSSe-based solar cells has recently reached 12.6% through the use of a CdS buffer layer and Se-rich CZTSSe with a bandgap in the 1.1–1.2 eV range.12 Despite this, the low open-circuit voltages (Vocs) of solar cells based on CZTSSe light absorbers, compared to those of CIGS and CdTe19,20 at the same bandgap energy, remain major issues. In particular, more states of deeper defect density are formed in CZTSSe at higher anionic mixing ratios (S/(S+Se) > 0.4), in addition to difficulties associated with S/(S+Se) fine tuning.10,21-23 With this in mind, an alternative method of altering the absorber bandgap is required, and one approach involves adjusting the metal cations during thin-film deposition. The bandgaps of Ge-incorporated CZTSe (CZTGSe) films can be readily tuned; hence, these are attractive materials for use with this alternative method. Several research groups have recently reported improvements in conversion efficiencies and Voc deficits in solar cells based on CZTGSe light absorbers.24-30 By controlling the Ge/(Sn+Ge) ratio (x), the CZTGSe bandgap can be tuned within the 1.0–1.4 eV range,28-32 and we achieved a conversion efficiency of 12.3% for a CZTGSe based solar cell with x = 0.2.29 Moreover, the Voc deficit in our device was 0.583 eV,


3


Page 4 of 41

which is 34 meV smaller than that of a champion kesterite device (0.617 eV).10 However, the cell performance of a heterojunction solar cell based on the kesterite absorber was much lower than that of a CIGS-heterojunction solar cell.33 Hence, further improvements in kesterite absorbers are important in order for them to catch up and surpass the cell performance of CIGSheterojunction solar cells. We consider that the band alignment at the n-type CdS/p-type CZTGSe heterointerface also plays a crucial role in realizing high-performance kesterite-based solar cells, as well as n-type CdS/p-type CIGS-heterojunction solar cells. It is well-known that, in addition to the film quality of the p-type CIGS absorber, band alignment at the interface between an n-type CdS-buffer and a p-type CIGS-absorber layer is important for achieving high-performance CdS/CIGSheterojunction solar cells.34,35 Consequently, if the conduction band offset (CBO) is negative, i.e., the position of the conduction band minimum (CBM) of CdS against the vacuum level is lower than that of CIGS at the interface, carrier recombination at trap states near the interface will be enhanced, resulting in poor cell performance. On the other hand, if the CBO has a small positive value, which means that the CBM of CdS against the vacuum level is higher than that of CIGS, carrier recombination at the interface will be suppressed, resulting in good cell performance. Although several theoretical17,36-41 and experimental studies42-51 on the electronic properties of CZTSSe surfaces and CdS/CZTSSe heterointerfaces have been reported, there are only a few reports32,52-56 on CdS/CZTGSe-heterointerface band alignment and the electronic properties of Ge-incorporated CZTSe surfaces, including (for example) defect formation. Hence, accumulating and understanding experimental CZTGSe and CdS/CZTGSe-heterointerface bandalignment data are crucial to achieving good cell-performance. In this study, we reveal band


4

Page 5 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


alignments at CdS/CZTGSe heterointerfaces with Ge/(Sn+Ge) ratios (x) in the 0–0.4 range, as well as the surface-electronic properties of CZTGSe using in-situ X-ray photoemission spectroscopy (XPS), ultra-violet photoemission spectroscopy (UPS), and inversed photoemission spectroscopy (IPES) techniques.

EXPERIMENTAL SECTION Experimental Setup To determine the band alignment at a CdS/CZTGSe heterointerface, the intrinsic electronic properties of the CdS and CZTGSe surfaces must be revealed in the absence of surface contamination, since valence band maximums (VBM)s, CBMs, and core levels are easily affected by surface contamination. We depict a schematic diagram of our experimental XPS, UPS, IPES, and CdS-deposition setup in Fig. 1. The various measurement chambers (XPS, UPS, IPES) and the CdS-deposition chamber are connected to the transfer chamber under ultrahigh vacuum (< ~1 × 10-7 Pa). Moreover, the Ar-circulating glovebox is connected using a load-lock (LL) chamber. Using this system, we performed XPS, UPS, and IPES and deposited CdS without exposure to air. XPS was conducted using a hemispherical electron-energy analyzer (ESCA SSX-100) with a monochromatic Al K X-ray source (h = 1486.6 eV) in order to determine the core level of each element in each sample. Valence band shapes were characterized by UPS using a hemispherical electron-energy analyzer (Scienta R-3000) with a monochromatic He I ultraviolet radiation source (h = 21.2 eV). The IPES system, which has a total energy resolution of 0.1 eV, was composed of an Erdman-Zipf type electron gun57 with a BaO cathode and a SrF2 window on a Hamamatsu R595 photomultiplier with a CuBe photocathode; the system was used to


5


Page 6 of 41

determine the shape of the conduction band. In this paper, we define the energy levels of the VBM and CBM relative to the Fermi level (EF) as VBMEF (= Ev – EF) and CBMEF (= EF – Ec), respectively, where Ev and Ec are the energy levels of the VBM and CBM relative to the vacuum level. It is worth noting that VBMEF < 0 and CBMEF > 0. For UPS, IPES, and XPS, EF is standard. VBMEF and CBMEF can be determined directly by UPS and IPES. The error tolerances of the band offsets in our analyses were ±0.15 eV.

Figure 1. Schematic diagram of the experimental XPS, UPS, IPES, and CdS-deposition setup under ultrahigh vacuum.

Sample Preparation CZTGSe films (thickness, 1.8 μm) were deposited on Mo-coated soda-lime glass substrates by the thermal co-evaporation of elemental Cu, Zn, Sn, Ge, and Se at a substrate temperature of


6

Page 7 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


200 °C.12,29 After deposition, the substrates were thermally annealed in a three zone furnace, after which they were etched using potassium cyanide (KCN) to remove excess Cu-related compounds, such as Cu2Se and CuSe. Following etching, the composition of the CZTGSe layer was characterized by X-ray fluorescence spectroscopy (SHIMADZU, EDX-720). We previously reported on the band alignment at a CdS/CZTSe (x = 0) heterointerface for an off-stoichiometric CZTSe absorber with a Cu/(Zn+Sn) ratio of ~0.843,44,48 because the cell performance of CdS/offstoichiometric-CZTSe-heterojunction solar cells is better than that of the CdS/stoichiometricCZTSe analogues. However, in the case of Ge-incorporated CZTSe, the performance of CdS/CZTGSe-heterojunction solar cells based on stoichiometric CZTGSe compositions {Cu/(Zn+Sn+Ge) ~ 1} is superior to that of off-stoichiometric CZTGSe. Therefore, to understand the effect of Ge incorporation into the CZTSe film on CdS/CZTGSe band alignment, we used stoichiometric CZTGSe with Cu/(Zn+Sn+Ge) ratios of ~1. Prior to XPS, UPS, IPES, and CdS deposition, the CZTGSe films were treated with aqueous NH3 solution in an Ar-gas circulating glovebox attached to the LL chamber, in order to remove adsorbed C and O on the CZTGSe surface. Surface contamination by adsorbed C and O will result in an inability to accurately determine VBMs, CBMs, and core levels. This surface treatment helps to reveal the intrinsic electronic properties of the CZTGSe surface without resorting to ion-beam irradiation. Following NH3 treatment, the samples were transferred from the glovebox into the LL chamber to remove adsorbed organic monomers and H2O from the CZTGSe surface by lamp annealing. The samples were then transferred from the LL chamber into the required analysis chamber or the CdS-deposition chamber via the transport chamber. CdS layers were sequentially deposited on the CZTGSe surfaces by thermal evaporation at room temperature. We acquired XPS, UPS, and IPES spectra in each CdS-deposition step (e.g. CdS


7


Page 8 of 41

film thickness 0, 3, 6, and 30 nm) via the transfer chamber without exposure to air until the CdS film on the CZTGSe substrate was 30 nm thick. Moreover, we also fabricated CdS/CZTGSe heterojunction solar cells in order to determine the bandgap energy and to confirm that our deposited CZTGSe films are effective as CdS/CZTGSe heterojunction solar cells. The following device structure was used: aluminum zinc oxide (AZO)/intrinsic zinc oxide (i-ZnO)/n-type CdS buffer/ CZTGSe absorber/Mo/soda-lime glass. AZO and i-ZnO were deposited by sputtering. The chemical-bath-deposition technique was used to deposit the CdS layer during the device-fabrication process,11,12,29 while CdS layers were deposited by thermal evaporation for determining band alignments at CdS/CZTGSe heterointerfaces.

THEORETICAL BACKGROUND In this study, we used the interface-induced band-bending (iibb) method to determine band alignments (CBOs and valence-band offsets (VBOs)) at CdS/CZTGSe heterointerfaces, which relies on the sum of the interface-induced band bendings of the overlayer and the bottom layer.47,49,51,58 The benefit of this method is that it accurately determines the band alignment at the interface between the overlayer (top layer) and the bottom layer, even if surface dipoles and/or secondary phases are generated at the heterointerface due to the intermixing of both the overlayer and the bottom layer. At least three layers are needed to determine the CBO and VBO using iibb: a bottom layer (layer I) devoid of an overlayer, a thin intermediate layer (layer II), which is grown on the bottom layer and facilitates determining the core level of each element in the intermediate and bottom layers (the UPS and IPES spectra originating between the bottom layer and the overlayer cannot be separated in this layer), and a top layer (layer III) that is a


8

Page 9 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


sufficiently thick overlayer. Moreover, the overlayer in layer III is needed to finish band bending, and the UPS and IPES signals only reflect on the overlayer.

Layer III (Top layer)

Layer II

Layer I

(Intermadiate layer)

(Bottom layer)

Vacuum Level Eiibb

Eiibb1 Eiibb2 CBM EF(CZTGSe)

CBM EF(CdS)

EF

IPES

CBO

IPES

UPS VBM EF(CZTGSe)

UPS VBO XPS

VBM EF(CdS) XPS

XPS E2(layerII) E2(layerIII)

E1(layerII)

Eiibb1

Eiibb2 Core level 2 Cd, S

XPS E1(layerI)

Core level 1 Cu, Zn, Sn, Ge, Se

CdS CZTGSe Interface

Figure 2. Depicting band alignment at a CdS/CZTGSe heterointerface as a function of EF, and the relationships between UPS, IPES, and XPS.

Fig. 2 shows the schematic diagram used to determine the band alignment at the interface of a CdS overlayer and a CZTGSe bottom layer relative to the Fermi-level standard under thermalequilibrium condition. The CBO and VBO at the interface between CdS and CZTGSe can be


9


Page 10 of 41

determined from Fig. 2 using the following equations:49

CBO = CBMEF(CdS) – CBMEF(CZTGSe) + Eiibb

(1),

VBO = VBMEF(CZTGSe) –VBMEF(CdS) – Eiibb

(2),

where CBMEF(CdS) and VBMEF(CdS) are the CBMEF and VBMEF of a sufficiently thick CdS overlayer on the CZTGSe bottom layer in layer III, while CBMEF(CZTGSe) and VBMEF(CZTGSe) are the CBMEF and VBMEF of the bottom layer surface in layer I. These energy levels in layers I and III can be determined by IPES and UPS. In contrast, Eiibb is the sum of the interface-induced changes due to band bending of both the thin CdS overlayer and the CZTGSe bottom layer, which can be determined from Fig. 2 using the following equation:58 Eiibb = Eiibb1 + Eiibb2

(3),

= {E1(layer II) ‐E1(layer I)} + {E2(layer III)‐E2(layer II)}

(4),

where Eiibb1 corresponds to the magnitude of the core-level shift of each element in the structure of the CZTGSe bottom layer before and after growth of the CdS thin film {Eiibb1 = E1(layer II) – E1(layer I)}; E1(layer I) is the core level of each element in the structure of the CZTGSe surface (the core levels of Cu, Zn, Sn, Ge, and Se) in layer I of Fig. 2, and E1(layer II) is the core level of each element in the structure of the CZTGSe bottom layer (the core levels of Cu, Zn, Sn, Ge, and Se) through the thin CdS overlayer in layer II of Fig. 2, and E2(layer II) is that of the CdS overlayer (Cd and S) in layer II of Fig. 2. In contrast, Eiibb2 corresponds to the magnitude of corelevel shift of each element in the structure of the CdS overlayer before and after CdS thin-film


10

Page 11 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


growth on the CZTGSe bottom layer {Eiibb2 = E2(layer III) – E2(layer II)}. E2(layer III) is the core level of each element (the core levels of Cd and S) in the structure of the sufficiently thick CdS overlayer on the CZTGSe bottom layer in layer III of Fig. 2, while E2(layer II) is the core level of each element (the core levels of Cd and S) in the structure of the thin CdS overlayer in layer II of Fig. 2. In this paper, we adopted a 3-nm-thick CdS overlayer on the CZTGSe bottom layer as the intermediate layer to determine Eiibb. This CdS thickness facilitates determining the magnitude of the core-level shift for each element in both the thin CdS overlayer and the CZTGSe bottom layer through the thin CdS overlayer by XPS, as shown in Fig. 2.

RESULTS AND DISCUSSION VBMEF and CBMEF in layer I. Fig. 3 shows the UPS (left) and IPES (right) spectra of the CZTGSe surface at Ge/(Sn+Ge) ratios (x) of 0, 0.2, and 0.4 relative to the Fermi level (EF). The VBM and CBM of CZTGSe were determined from the points where the linear extrapolations (red lines) of the leading edges of the UPS and IPES spectra cross the baseline, respectively. The determined VBM and CBM energy levels relative to the Fermi level {VBMEF(CZTGSe) and CBMEF(CZTGSe)} and bandgap energies of CZTGSe {Eg(CZTGSe) = | VBMEF(CZTGSe)| + | CBMEF(CZTGSe)|} are summarized in Table. 1.


11


UPS ex. He (I) EF

(b)

IPES EF

CZ(T0.6G0.4)Se

CZ(T0.8G0.2)Se

CBMEF

VBMEF

CZTSe

-6

-5

-4

-3

-2

VBMEF

CBMEF

VBMEF

CBMEF

-1

0

0

1

2

3

4

5

Electron Energy Relative to the Fermi level (eV)

UPS ex. He (I)

Normalized Intensity (arb. units)

(a) Normalized Intensity (arb. units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 41

6

-2

EF

VBMEF

VBMEF

VBMEF

-1

0

Electron Energy Relative to the Fermi Level (eV)

Figure 3. (a) UPS (left) and IPES (right) spectra of the Cu2ZnSnSe4 (x = 0), Cu2Zn(Sn0.8Ge0.2)Se4 (x = 0.2), and Cu2Zn(Sn0.6Ge0.4)Se4 (x = 0.4) surfaces as functions of EF. (b) Enlarged UPS spectra from Fig. 3(a) near EF relative to EF.

Table 1. IPES- and UPS-determined VBMEF, CBMEF, and Eg values (eV) for CZTGSe surfaces with Ge/(Sn+Ge) ratios (x) of 0, 0.2, and 0.4. These data correspond to the CZTGSe bottom layer in layer I of Fig. 2. Bottom layer (layer I) CZTSe (x = 0) CZ(TG)Se (x = 0.2) CZ(TG)Se (x = 0.4)

VBMEF(CZTGSe)

CBMEF(CZTGSe)

Eg(CZTGSe)

-0.44

+0.59

1.03

-0.45

+0.64

1.09

-0.42

+0.78

1.20


12

Page 13 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1 reveals that only the CBMs of the CZTGSe shift toward higher energies with increasing Ge/(Sn+Ge) ratio, whereas the VBMs remain constant; this observation is further discussed below. Moreover, we noticed that the EFs at the CZTGSe surface are positioned near the center of the band gap (Table l and Fig. 3). In the case of VBMEF (CZTGSe), which ranges between -0.45 and -0.42 eV, the hole concentration (nh) is only of the order of ~1011 cm-3, and was calculated by: nh = Nv  exp{–(EF – Ev)/(kT)}

(5),

where Nv is the effective density of state of the valence band, and -(EF – Ev) corresponds to the VBMEF. In this calculation, Nv is assumed to be 2.4 × 1018 cm-3 at room temperature.59 The calculated carrier concentration is significantly lower than that determined from capacitancevoltage (C-V) experiments, despite carrier concentrations of ~1016 cm-3 and the p-type nature of our CZTGSe samples.60 It is well-known that the UPS technique is very sensitive to surface electronic properties compared to the C-V technique; consequently, the UPS-determined carrier concentration of 1011 cm-3 mainly reflects the electronic nature of the surface. Hence, we conclude that the electronic properties of the surface and bulk are different. There are two possible reasons for the low carrier concentration of 1011 cm-3 determined by UPS. One is strong Fermi-level pinning at the surface. The other is hole deficiencies that are concentrated near the surface, which we believe to be the main reason and is further discussed below (Surface Electronic Properties). The Eg values determined by UPS and IPES for CZTGSe with Ge/(Sn+Ge) ratios of 0, 0.2, and 0.4 are consistent with those determined by external quantum efficiency (EQE) spectra. Fig. 4(a) displays the EQE spectra of CdS/CZTGSe heterojunction solar cells, while Fig. 4(b) shows the plots used to determine bandgaps, which were obtained by extrapolating the linear region of the


13


relationship between [hν × ln(1 – EQE)]2 and photon energy (hν) near the band edge of each material.12,61,62 The bandgap energy increases with increasing x, as shown in Fig. 4(b); Eg values for CZTGSe absorbers with x values of 0, 0.2, and 0.4 were found to be 0.97, 1.11, and 1.22 eV, respectively. 1 (a)

CZ(T0.6G0.4)Se CZ(T0.8G0.2)Se CZTSe (x=0)

(b)

{E x ln(1–EQE)}

2

100

EQE Intensity (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 41

50

Eg (x=0)

Eg (x=0.2)

Eg (x=0.4)

0

0 500

1000 Wavelength (nm)

1500 0.9

1.0 1.1 1.2 Photon Energy (eV)

1.3

Figure 4. (a) EQE spectra and (b) relationships between [hν ln(1-EQE)]2 and hν for CdS/CZTGSe heterojunction solar cells with Ge compositions (x) of (●) 0, (○) 0.2, and (□) 0.4.

VBMEF(CdS) and CBMEF(CdS) in layer III. As mentioned above, a sufficiently thick CdS overlayer is needed to determine the VBMEF and CBMEF values of the CdS overlayer corresponding to layer III in Fig. 2. From experimental UPS and IPES results, we found that a 30-nm-thick CdS overlayer is sufficient to determine the VBMEF and CBMEF of the CdS overlayer on the CZTGSe. Fig. 5 shows representative UPS (left) and IPES (right) spectra relative to EF for (a) 0-, 3-, 6-, and 30-nm-thick CdS overlayers on CZTGSe with a Ge composition (x) of 0.2; “0-nm-thick” corresponds to the CZTGSe surface.


14

Page 15 of 41

Enlargements of the UPS spectra displayed in Fig. 5(a) are shown in Fig. 5(b), as functions of EF.

-6

(b)

IPES

UPS ex. He(I) EF

EF

VBM EF

CdS 30 nm

Normalized Intensity (arb. units)

(a) Normalized Intensity (arb. units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


CBM EF CdS 6 nm

CdS 3 nm

CdS 0 nm

-5

-4

-3

-2

VBM EF

CBM EF

-1

0

0

1

2

3

4

5

Electron Energy Relative to the Fermi level (eV)

6

UPS ex. He(I) EF VBM EF

VBM EF

-3

-2

-1

0

Electron Energy Relative to the Fermi Level (eV)

Figure 5. UPS(left) and IPES(right) spectra relative to the Fermi level for (a) 0-, 3-, 6-, and 30nm-thick CdS overlayers on CZTGSe with a Ge composition (x) of 0.2. (b) Enlarged UPS spectra from Fig. 5(a) near EF relative to EF.

IPES signals originating from the anti-bonding Sn 5s, Ge 4s and Se 4p orbitals37,47,53,63 appear at around 2 eV for the CZ(T0.8G0.2)Se surface; this peak rapidly decreases in intensity with increasing CdS film thickness and is not observed when the CdS overlayer is 30 nm thick. In contrast, the UPS signal originating from Cu 3d and Se 4s anti-bonding orbitals37,47,53,63 appears at around -3 eV for the CZ(T0.8G0.2)Se surface.47,64 The UPS peak for the Cu 3d and Se 4p hybrid orbital is completely absent when the CdS overlayer is 30 nm thick, and other UPS peaks of


15


Page 16 of 41

different origin appeared at -3.5 eV, which are ascribable to Cd 5s and S 3p anti-bonding orbitals.64 Hence, we conclude that the UPS and IPES signals acquired when the CdS overlayer was 30 nm thick are not derived from mixed bottom and overlayer signals, but correspond only the CdS overlayer, which acts as layer III in Fig. 2. Moreover, the bandgap energy of the 30-nm-thick CdS overlayer on CZ(T0.8G0.2)Se was determined by UPS and IPES to be 2.50 eV, which corresponds to the typical band gap energy of CdS64-67. Furthermore, the bandgap energy and the position of the core Cd 3d level remain constant when the CdS overlayer is more than 30 nm thick. These results indicate the end of band bending by CdS growth on the CZTGSe bottom layer. With this in mind, we conclude that a 30-nm-thick CdS overlayer is sufficient for use as layer III because it fulfills the requirements of layer III. This CdS-growth behavior on CZ(T0.8G0.2)Se is similar to that on CZTGSe with Ge/(Sn+Ge) ratios 0 and 0.4; consequently, we determined VBMEF, CBMEF, and Eg values for 30-nm-thick CdS overlayers on CZTGSe with other Ge/(Sn+Ge) ratios using the same procedures, the results of which are summarized in Table 2. We consider that the electronic properties of 30-nm-thick CdS can be treated as those of the bulk because the Eg(CdS) values are also independent of the Ge composition x in CZTGSe, as shown in Table 2. Moreover, we consider that the fundamental electronic properties of the CdS deposited by the evaporation technique (MBE-CdS) and by CBD (CBD-CdS) are basically the same because the bandgap energy (Eg) and the position of the Fermi level in CBD-CdS69 determined by UPS and IPES are consistent with those of our prepared MBE-CdS.

Table 2. IPES and UPS-evaluated VBMEF, CBMEF, and Eg values (eV) for 30-nm-thick CdS overlayers on CZTGSe substrates with Ge compositions (x) of 0, 0.2, and 0.4.


16

Page 17 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Bottom layer (layer II) 30-nm CdS on CZTSe (x = 0) 30-nm CdS on CZ(TG)Se (x = 0.2) 30-nm CdS on CZ(TG)Se (x = 0.4)

VBMEF(CdS)

CBMEF(CdS)

Eg(CdS)

-2.15

+0.33

2.48

-2.17

+0.33

2.50

-2.15

+0.33

2.48

Interface-induced band bending in layer II. Eiibb (= Eiibb1 + Eiibb2) can be calculated using eq. (4), and each core-level term (in layers I, II, and III) can be evaluated by XPS, as shown in Fig. 2. Fig. 6 shows representative Cu 2p3/2,　Zn 2p3/2, Sn 3d5/2, Ge 3d, Se 3d, and Cd 3d5/2 XPS spectra acquired through the CdS overlayer for 0-, 3-, 6-, and 30-nm-thick CdS overlayers, respectively, on a CZTGSe bottom layer with a Ge/(Sn+Ge) ratio of 0.2. Each elemental XPS signal was fitted using Gaussian-Lorentzian functions after background correction by the Shirley method; Eiibb values for Ge/(Sn+Ge) ratios of 0, 0.2, and 0.4 were determined from these spectra and are summarized in Table 3.


17


60 (a)

(b)

Cu 2p 3/2

Zn 2p 3/2

Intensity (cps)

Intensity (cps)

40 40

20

0

20

0

936

934 932 930 Binding Energy (eV)

1024

928

1022 1020 Binding Energy (eV)

1018

4

(c)

(d)

Sn 3d 5/2

Intensity (cps)

Intensity (cps)

40

20

Ge 3d

2

0 0

490

36


30

(e)

400

Se 3d

20

10

0

60

(f)


28

Cd 3d Cd 3d5/2

Intensity (cps)

Intensity (cps)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 41

Cd 3d3/2 200

0

55 Binding Energy (eV)

50

415

410 405 Binding Energy (eV)

Figure 6. Representative (a) Cu 2p3/2, (b) Zn 2p3/2, (c) Sn 3d5/2, (d) Ge 3d, (e) Se 3d, and (f) Cd 3d (3d5/2, 3d3/2) XPS spectra of a CZ(T0.8G0.2)Se bottom layer acquired through CdS overlayers of different thickness: black, 0 nm; red, 3 nm; blue, 6 nm; and green, 30 nm.


18

Page 19 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 3. Eiibb values determined from the core-level shifts observed in the XPS spectra of the CZTGSe bottom layer, with Ge/(Ge+Sn) ratios of 0, 0.2, and 0.4, through a 3-nm-thick CdS overlayer (as layer II in Fig. 2).

Bottom layer Core level shift (eV)

CZTSe (x = 0)

CZTGSe (x = 0.2)

CZTGSe (x = 0.4)

Eiibb1(Cu 2p3/2) Eiibb1(Zn 2p3/2) Eiibb1(Sn 3d5/2) Eiibb1 Eiibb1(Ge 3d) Eiibb1(Se 3d) Eiibb1(ave) Eiibb2(Cd 3d5/2) Eiibb2 Eiibb2(Cd 3d5/2) Eiibb2(ave) Eiibb(ave) {= Eiibb1(ave)+ Eiibb2(ave)}

+0.06 +0.06 +0.12 +0.13 +0.09 +0.53 +0.53 +0.53 +0.62

+0.04 +0.09 +0.15 +0.07 +0.07 +0.08 +0.53 +0.53 +0.53 +0.61

+0.30 +0.24 –0.14 +0.18 +0.15 +0.50 +0.50 +0.50 +0.65

In this paper, we used average values of Eiibb {Eiibb(ave)} that were determined by averaging Eiibb1 {Eiibb1(ave)} and Eiibb2 {Eiibb2(ave)}, which enhance the accuracies of the determined bandbending magnitudes, as shown in layer II of Fig. 2. We did not use Eiibb(Sn) to determine Eiibb1(ave) for CZ(T1-xGx) when x = 0.4 because the signal was very weak and noisy. Moreover, we did not use the S core-level shift to determine iibb when 0 ≤ x ≤ 0.4 because the S 2p and Se 3p binding-energy signals were close and overlapped when the CdS overlayer was deposited on the CZTGSe substrate. As a result, it was difficult to separate these signals. The Eiibb1(ave) and Eiibb2(ave) values for CZTGSe (x ≤ 0.4) have positive values, as shown in Table 3. These results suggest that upward bending and downward bending toward the CdS/CZTGSe interface occurs at the CdS overlayer and CZTGSe bottom layer, respectively, as shown in layer II of Fig. 2. Moreover, we propose that the Fermi level at the CZTGSe surface is not pinned because we


19


Page 20 of 41

observe core-level shifts for each element in CZTGSe with only a 3-nm-thick CdS film grown on the CZTGSe, as shown by the Eiibb1(ave) data listed in Table 3. This means that the shift in the Fermi level at the CZTGSe surface is facilitated by a small amount of compensating charge. These observations are further discussed below (Surface Electronic Properties).

CBOs and VBOs. We determined the CBOs and VBOs at CdS/CZTGSe heterointerfaces with Ge compositions (x) of 0, 0.2, and 0.4, respectively. Each term in eqs. (1) and (2) are listed in Tables 1–3; the band alignments of CdS/CZTGSe (x = 0, 0.2, and 0.4) using these evaluated terms are shown schematically in Fig. 7. We found that the CBOs depended strongly on composition; CBOs were determined to be +0.36, +0.30, and +0.20 eV, for x = 0, 0.2, and 0.4, respectively. CBO was observed to decrease with increasing x. A positive CBO value indicates that the CBM position of CdS is higher than that of CZTGSe, which is so-called “spike” conduction band alignment. In contrast, the VBOs of CdS/CZTGSe with x = 0, 0.2, and 0.4 are +1.09, +1.11, and +1.08 eV, respectively; the VBO is clearly independent of Ge composition. (a)

CBO +0.36 eV

(b)

CBO +0.30 eV

CBMEF(CZTSe)

CBMEF(CdS)

+0.59 eV

+0.33 eV

(c)

CBMEF(CdS)

+0.64 eV

+0.33 eV

EF

-0.44 eV

EF

-0.45 eV

-2.17 eV

Interface

CZTSe (x=0)

+0.78 eV

EF

-0.42 eV VBM EF(CZTGSe)

-2.15 eV

VBO +1.08 eV

VBO +1.11 eV

VBM EF(CdS)

VBM EF(CdS)

CdS

CBMEF(CdS)

VBM EF(CZTGSe)

VBO +1.09 eV

VBM EF(CdS)

CBMEF(CZTGSe)

+0.33 eV

VBM EF(CZTSe) -2.15 eV

CBO +0.20 eV

CBMEF(CZTGSe)

CdS

Interface

CZTGSe (x=0.2)

CdS

Interface

CZTGSe (x=0.4)

Figure 7. Schematic diagram of the band alignment at the CdS/CZTGSe interface with Ge/(Sn+Ge) ratios (x) of (a) 0, (b) 0.2, and (c) 0.4.


20

Page 21 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In order to more deeply understand the dependencies of CBO and VBO on Ge composition (x), we also determined CBOs and VBOs for several CdS/CZTGSe heterointerface samples by the same procedures. Fig. 8 shows CBOs and VBOs as functions of x below 0.4. Solid circles ( ●) represent the CBOs and VBOs for CdS/stoichiometric-CZTGSe heterointerfaces with a Cu/(Zn+Sn+Ge) ratio of ~1 in this work. The open circles (○) are the CBO and VBO values for a CdS/off-stoichiometric-CZTSe with a Cu/(Zn+Sn) ratio of ~0.8 from our previous work.31 It is well known that the cell performance of a Cu-poor and Zn-rich off-stoichiometric-CZTSe absorber is superior to that of a stoichiometric CZTSe absorber in a CZTSe solar cell. In contrast, the cell performance of a nearly stoichiometric-CZTGSe absorber is better in off-stoichiometricCZTGSe solar cell.12,29 Therefore, we determined the CBOs and VBOs for heterostructures using near stoichiometric CZTGSe with a Cu/(Zn+Sn+Ge) ~ 1.


21


CBO (eV)

0.6

(a)

0.4

0.2

0.0 1.5 (b)

VBO (eV)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 41

1.0

0.5

0.0 0

0.1 0.2 0.3 0.4 Ge/(Sn+Ge) ratio (x)

Figure 8 (a) CBOs and (b) VBOs as functions of Ge/(Ge+Sn) ratio (x) below 0.4; (●) and (○) indicate stoichiometric and off-stoichiometric CZTGSe with Cu/(Zn+Sn+Ge) ratios of ~1 and ~0.8, respectively.

We next discuss the relationship between CBO, VBO, and Ge composition using the stoichiometric CZTGSe. The CBOs and VBOs are expressed by following equations:

VBO = VBMvac(CdS) – VBMvac(CZTGSe)

(6),

CBO = CBMvac(CZTGSe) – CBMvac(CdS)

(7),


22

Page 23 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


where CBMvac(CdS), CBMvac(CZTGSe), VBMvac(CdS), and VBMvac(CZTGSe) are the CBM and VBM energy levels of CdS and CZTGSe relative to the vacuum level, respectively. The constant energy shift of VBMvac(CZTGSe) with x is the same as that of the VBO because VBMvac(CdS) corresponds to the ionization potential and is constant due to the characteristics of CdS, as expressed by eq. (6). The VBMs for CZTGSe are mainly structured by the occupied antibonding Cu 3d and Se 4p orbital,53 and Cu-Se is a weak covalent bond. The Ge 4s and Sn 5s orbitals do not strongly contribute to chemical bonding due to the inert-pair effect. Moreover, we do not need to consider the contributions of the Zn 3d, Se 4p, Zn 4s, or Se 4p bonding orbitals to the VBM because they are localized at deeper energies than that of the VBM structured by the occupied Cu 3d and Se 4p antibonding orbitals. Hence, the VBOs remain constant even when the Sn in CZTGSe is substituted by Ge because the VBM, and the Cu and Se chemical bonds, are largely unchanged. In contrast, the energy shift of CBMvac(CZTGSe) at different x values is equal to that of the CBO because CBMvac(CdS), which corresponds to electron affinity and is constant due the characteristic of CdS, is expressed by eq. (7). The CBMs of CZTGSe are structured by the unoccupied antibonding Sn 5s, Ge 4s, and Se 4p orbitals;55 hence, the energy levels of CBMvac(CZTGSe) depend on Sn-Se and Ge-Se chemical bonds. Concerning the CBM structured by anti-bonding orbitals, CBMvac(CZTGSe) for short bond lengths are closer to the vacuum level compared to those with long bond lengths. Actually, our group confirmed that Eg(CZTGSe) increases linearly with increasing Ge composition (x), and the CZTGSe lattice constant decreases linearly with increasing x. Here, the energy positions of the CBM and VBM could be influenced by the bond length.53,73 The lattice constant following substitution of Sn for Ge in


23


Page 24 of 41

CZTSe is about 2% shorter than that prior to substitution, while the Ge-Se bond length is about 7% shorter than the Sn-Se bond length, despite the Cu-Se bond length following substitution of Sn for Ge being only 0.3% shorter than that of prior to substitution, both experimentally and computationally.53 The change in the IV(Sn, Ge)-Se bond length is significantly larger than that of the Cu-Se bond length upon substitution of Sn for Ge in CZTSe. As a result, the magnitude of the shift in the CBM energy with increasing Ge composition x is expected to be larger than that of the VBM. We conclude that the experimental and calculational results are consistent with our evaluated VBM and CBM as functions of Ge composition x, as shown in Fig. 8(a) and (b), while we cannot clearly observe the shift in the VBM energy by Ge substitution for 0 ≤ x ≤ 0.4 because it is too small to detect under our experimental conditions. Moreover, the trend of the VBMvac and CBMvac values for CZTGSe can be also understood by analogy with the Cu(In1-zGaz)Se2 alloy system73,74 because the CBM consists of In-Ge and Se, while the VBM consists of Cu and Se. The CBM shifts due to the Ga and In alloy. The CBM of CZTGSe, which consists of IV-Se is expected to shift in an IV alloy system, by analogy with CIGSe. Concerning them, we conclude that our experimental results (Fig. 8) are consistent with VBM and CBM formation based on theoretical considerations. We next discuss the differences in CBO and VBO between CdS/off-stoichiometric CZTSe and CdS/stoichiometric CZTSe, as shown in Fig. 9. In order to easily understand these CBO and VBO differences, a schematic band diagram relative to the vacuum level is shown in Fig. 9.


24

Page 25 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Vacuum level

~

~

CBMvac (CdS)

CBO +0.36 eV

CBMvac (CZTSe)

CBO +0.56 eV CBMvac (CZTSe)

VBM vac (CZTSe) VBM vac (CZTSe)

VBO +1.09 eV

VBM vac (CdS)

CdS

VBO +0.89 eV

stoichiometric

off-stoichiometric

CZTSe(x=0)

CZTSe(x=0)

Figure 9. Schematic energy-level diagram for CdS, off-stoichiometric, and stoichiometric CZTSe relative to the vacuum level.

The CBO of the off-stoichiometric CZTSe is +0.20 eV larger than that of the stoichiometric CZTSe. Moreover, the VBO of the off-stoichiometric CZTSe is -0.20 eV smaller than that of stoichiometric CZTSe, as shown in Fig. 9. These differences in CBO and VBO energy are relatively large when compared to the magnitude of the bandgap energy shift (ΔEg) associated with moving from CZTSe (x = 0) to CZTGSe (x = 0.4), since ΔEg is 0.17 eV (Table 1). On the basis of this analysis, we consider that phase separation possibly occurs in the off-stoichiometric CZTSe due to the large Cu-poor and Zn-rich off-stoichiometric Cu/(Zn+Sn+Ge) composition of ~0.8. Actually, the Sn/Zn solid-solution region is quite narrow;70 Thus further investigations are required in order to better understand the phase separation and electronic properties of the off-


25


Page 26 of 41

stoichiometric-CZTSe surface and CdS/off-stoichiometric CZTSe heterointerface.

Surface electronic properties Based on the CZTGSe bending and the values of Eiibb1(ave) from our experiments, we conclude that the Fermi-level is not pinned at the CZTGSe surface. In order to evaluate defect states near the surface, we measured core-level shifts for each element when irradiated with a continuous-wave (cw) laser. Moreover, in order to understand the depth distribution of the defect states near the surface, we measured the wavelength dependence of the core-level shift for each element using three cw lasers operating at wavelengths of 405, 685, and 938 nm; these laser wavelengths penetrate approximately 40, 160, and 390 nm, respectively, as determined from the CZTSe optical constants.16, 71 Fig. 10(a) shows average core-level shifts (from the dark) of each element upon irradiation of surfaces with Ge compositions (x) of 0, 0.2, and 0.4 with ~100 mW/cm2 (~1 sun) lasers at the wavelengths mentioned above, while Fig. 10(b) shows a schematic band diagram of the free CZTGSe surface under light and dark conditions. In this section, we define the VBMEF(CZTGSe) relative to the Fermi level at the surface and in the bulk as VBMEF(Surf) and VBMEF(Bulk), respectively. Moreover, the magnitude of CZTGSe band bending between the surface and the bulk (Ebb) is defined diagrammatically in Fig. 10(b). The black solid lines show the VBM, CBM, and the core level of CZTGSe under dark conditions, while the blue, orange, and red dotted lines show the VBM, CBM, and core level of CZTGSe when irradiated at laser wavelengths of 405, 685, and 938 nm, respectively. The zeros in Fig. 10(a) and (b) represent the core levels of each element at the surface under dark conditions.


26

Page 27 of 41

0.35

Averaged core level shift (eV)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

CZTSe (x=0) CZTGSe (x=0.2) CZTGSe (x=0.4)

0.30 0.25

(b)

0.20 0.15 0.10 0.05 0.00 -0.05

400

600

800

1000

Excitation wavelength (nm)

Figure 10. (a) Average core level shifts upon laser irradiation relative to those under dark condition for CZTGSe with Ge compositions (x) of (●) 0, (○) 0.2, and (□) 0.4, respectively, as functions of excitation wavelength. (b) Schematic band diagram of the CZTGSe surface under dark conditions and when laser irradiated.

Clear core-level shifts were observed upon laser irradiation, as shown in Fig. 10(a). Core-level shifts at Ge compositions (x) of 0, 0.2, and 0.4 at a wavelength of 405 nm, which is the most sensitive, were found to be +0.11, +0.12, and +0.27 eV, respectively. These shifts are comparable to the band bending Ebb observed in Fig. 10(a) under 1-sun conditions, which indicates that Fermi-level is not pinned at the free CZTGSe surface. The Ebb {=VBMEF(Surf) VBMEF(Bulk)} in Fig. 10(b) can be determined by the difference between VBMEF(Surf) and VBMEF(Bulk); VBMEF(Surf) was evaluated as shown in Table. 1. VBMEF(bulk) at a hole concentration in the bulk of 1016 cm-3 was determined to be 0.18 eV using eq. (5) and an Nv of 2.4 × 1018 cm-3; therefore, the Ebb under dark conditions is calculated to be 0.24 eV at a hole


27


Page 28 of 41

concentration in the bulk of 1016 cm-3. Moreover, we found that the hole-deficient layer is distributed near the CZTGSe surface rather than in the bulk on the basis of the laser-wavelength dependence of the core-level shift. As shown in Fig. 10(a), the average core-level shift increases with decreasing laser wavelength, which indicates that larger shifts were observed when excited closer to the surface; i.e., the holedeficient region is located closer to the surface because more band bending occurs in more holedeficient regions. We consider that these defect states near the surface region are relatively low in energy and can be sufficiently filled by photo-induced carriers under only one-sun illumination conditions because the Fermi level shifts under one-sun illumination conditions, which are not strong excitation conditions and do not generate many compensating charges. Hence, we conclude that the amount of defect states is relatively small and that the Fermi level at the CZTGSe surface is not pinned by surface defect states. We speculate that the origin of this hole-deficient layer is related to the formation of donors rather than a reduction in the number of accepters, and one such possible donor state is the ZnCuantisite defect, which is reported to be a shallow donor state.41,72 Actually, there tends to be less Cu near the surface compared the bulk following annealing in our experiments; therefore, ZnCu antisites are easily formed at the surface. The origin of the hole-deficient layer at the CZTGSe surface remains unclear at present, however, these defect states could affect solar-cell performance. Further investigations are therefore needed in order to realize high solar-cell performance based on kesterite.

CONCLUSIONS We evaluated the band alignment at CdS/CZTGSe heterointerfaces with Ge/(Sn+Ge) ratios (x)


28

Page 29 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in the 0–0.4 range by in-situ XPS, UPS, and IPES. The CBO at the CdS/CZTGSe heterointerface decreased linearly from +0.36 to +0.20 eV with increasing Ge composition. These CBOs mean spike conduction band alignments and favorable conditions for suppressing carrier recombination at the interface. In contrast, the VBO at the heterointerface was independent of Ge composition and remained constant around +1.09 eV. We showed CBOs and VBOs determined by the iibb method involving relationships with Ge composition are consistent with those obtained from first-principles calculations. Moreover, both UPS and IPES revealed that the Fermi levels at the CZTGSe surface are located near the center of the bandgap where the hole concentration is of the order of 1011 cm-3; this hole concentration is much smaller than that (~1016 cm-3) of the CZTGSe bulk. In particular, laser-irradiated XPS revealed that the origin of the hole deficiency near the surface is due to defects distributed closer to the CZTGSe surface than in the bulk. Furthermore, we showed that a small amount of these defect states near the CZTGSe surface are sufficiently filled by photoinduced carriers under only one-sun illumination conditions, or by compensating charges due to the very thin CdS deposition on the CZTGSe. Hence, we conclude that the Fermi level is not pinned at the CZTGSe surface. While the origins of these hole-deficient states at the CZTGSe surface are presently unclear, this hole-deficient layer could play an important role in determining the performance of kesterite-based solar cells. Therefore, further investigations are needed in order to understand this phenomenon.

ASSOCIATED CONTENT Supporting Information. Representative XRF spectrum of CZTGSe with x = 0.2; Representative survey XPS spectra of


29


Page 30 of 41

CZTGSe (x = 0.2) with 0–30-nm thick CdS overlayers; S 2p and Se 3p XPS signals as functions of CdS film thickness on the CZTGSe substrate (PDF). This Supporting Information is available free of charge on the ACS Publications website at DOI: AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] Author Contributions All authors have given approval to the final version of the manuscript. ORCID Takehiko Nagai: 0000-0002-1873-5764 Shinho Kim: 0000-0002-8895-0232 Hitoshi Tampo: 0000-0002-6666-0285 Hajime Shibata: 0000-0002-2405-2327 Koji Matsubara: 0000-0002-6707-5010 Shigeru Niki: 0000-0002-3877-2028 Notes The authors declare no competing financial interests. ACKNOWLEDGMENT The authors thank M. Iioka, H. Higuchi, and S. Takaesu for their assistance with experiments and their technical support. This research was supported in part by the New Energy and Industrial Technology Development Organization (NEDO) under the Ministry of Economy,


30

Page 31 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Trade and Industry (METI). REFERENCES (1) Haegel, N. M.; Margolis, R.; Buonassisi, T.; Feldman, D.; Froitzheim, A.; Garabedian, R.; Green, M.; Glunz, S.; Henning, H.-M.; Holder, B.; Kaizuka, I; Kroposki, B.; Matsubara, K.; Niki, S.; Sakurai, K.; Schindler, R. A.; Tumas, W.; Weber, E. R.; Wilson, G.; Woodhouse, M.; Kurtz, S. Terawatt-Scale Photovoltaics: Trajectories and Challenges. Science 2017, 356, 141– 143. (2) Wadia, C.; Alivisatos, A. P.; Kammen, D. M. Materials Availability Expands the Opportunity for Large-Scale Photovoltaics Deployment. Environ. Sci. Technol. 2009, 43, 2072–2077. (3) Green, M. A. Consolidation of Thin-Film Photovoltaic Technology: The Coming Decade of Opportunity. Prog. Photovoltaics Res. Appl. 2006, 14, 383–392. (4) Andersson, B. A. Materials Availability for Large-Scale Thin-Film Photovoltaics. Prog. Photovoltaics Res. Appl. 2000, 8, 61–76. (5) Ito, K.; Nakazawa, T. Electrical and Optical Properties of Stannite-Type Quaternary Semiconductor Thin Films. Jpn. J. Appl. Phys. 1988, 27, 2094–2097. (6) Katagiri, H. Cu2ZnSnS4 Thin Film Solar Cells. Thin Solid Films 2005, 480–481, 426–432. (7) Zoppi, G.; Forbes, I.; Miles, R. W.; Dale, P. J.; Scragg, J. J.; Peter, L. M. Cu2ZnSnSe4 Thin Film Solar Cells Produced by Selenisation of Magnetron Sputtered Precursors. Prog. Photovoltaics Res. Appl. 2009, 17, 315–319. (8) Katagiri, H.; Jimbo, K.; Maw, W. S.; Oishi, K.; Yamazaki, M.; Araki, H.; Takeuchi, A. Development of CZTS-Based Thin Film Solar Cells. Thin Solid Films 2009, 517, 2455–2460.


31


Page 32 of 41

(9) Todorov, T. K.; Reuter, K. B.; Mitzi, D. B. High-Efficiency Solar Cell with Earth-Abundant Liquid-Processed Absorber. Adv. Mater. 2010, 22, E156–E159. (10) Wang, W.; Winkler, M. T.; Gunawan, O.; Gokmen, T.; Todorov, T. K.; Zhu, Y.; Mitzi, D. B. Device Characteristics of CZTSSe Thin-Film Solar Cells with 12.6% Efficiency. Adv. Energy Mater. 2013, 4 (7), 1301465. (11) Kim, K. M.; Liao, K. H.; Tampo, H.; Shibata, H.; Niki, S. Cu2ZnSnSe4 Thin-Film Solar Cells Fabricated Using Cu2SnSe3 and ZnSe Bilayers. Appl. Phys. Express 2015, 8, 042301. (12) Kim, S.; Kim, K. M.; Tampo, H.; Shibata, H.; Niki, S. Improvement of Voltage Deficit of Ge-Incorporated Kesterite Solar Cell with 12.3% Conversion Efficiency. Appl. Phys. Express 2016, 9, 102301. (13) Katagiri, H.; Sasaguchi, N.; Hando, S.; Hoshino, S.; Ohashi, J.; Yokota, T. Preparation and Evaluation of Cu2ZnSnS4 Thin Films by Sulfurization of E–B Evaporated Precursors. Sol. Energy Mater. Sol. Cells 1997, 49, 407–414. (14) Wibowo, R. A.; Kim, W. S.; Lee, E. S.; Munir, B.; Kim, K. H. Single Step Preparation of Quaternary Thin Films by RF Magnetron Sputtering from Binary Chalcogenide Targets. J. Phys. Chem. Solids 2007, 68, 1908–1913. (15) Liu, F.; Li, Y.; Zhang, K.; Wang, B.; Yan, C.; Lai, Y.; Zhang, Z.; Li, J.; Liu, Y. In Situ Growth of Cu2ZnSnS4 Thin Films by Reactive Magnetron Co-Sputtering. Sol. Energy Mater. Sol. Cells 2010, 94 (12), 2431–2434. (16) Hirate, Y.; Tampo, H.; Minoura, S.; Kadowaki, H.; Nakane, A.; Kim, K. M.; Shibata, H.; Niki, S.; Fujiwara, H. Dielectric Functions of Cu2ZnSnSe4 and Cu2SnSe3 Semiconductors. J.


32

Page 33 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Appl. Phys. 2015, 117, 015702. (17) Chen, S.; Gong, X. G.; Walsh, A.; Wei, S.-H. Crystal and Electronic Band Structure of Cu2ZnSnX4 (X=S and Se) Photovoltaic Absorbers: First-Principles Insights. Appl. Phys. Lett. 2009, 94, 041903. (18) Salomé, P. M. P.; Malaquias, J.; Fernandes, P. A.; Ferreira, M. S.; da Cunha, A. F.; Leitão, J. P.; González, J. C.; Matinaga, F. M. Growth and Characterization of Cu2ZnSn(S,Se)4 Thin Films for Solar Cells. Sol. Energy Mater. Sol. Cells 2012, 101, 147–153. (19) Kaur, K.; Kumar, N.; Kumar, M. Strategic Review of Interface Carrier Recombination in Earth Abundant Cu–Zn–Sn–S–Se Solar Cells: Current Challenges and Future Prospects. J. Mater. Chem. A 2017, 5, 3069–3090. (20) Kim, J.; Shin, B. Strategies to Reduce the Open-Circuit Voltage Deficit in Cu2ZnSn(S,Se)4 Thin Film Solar Cells. Electron. Mater. Lett. 2017, 13, 373–392. (21) Todorov, T. K.; Tang, J.; Bag, S.; Gunawan, O.; Gokmen, T.; Zhu, Y.; Mitzi, D. B. Beyond 11% Efficiency: Characteristics of State-of-the-Art Cu2ZnSn(S,Se)4 Solar Cells. Adv. Energy Mater. 2013, 3, 34–38. (22) Duan, H.-S.; Yang, W.; Bob, B.; Hsu, C.-J.; Lei, B.; Yang, Y. The Role of Sulfur in Solution-Processed Cu2ZnSn(S,Se)4 and Its Effect on Defect Properties. Adv. Funct. Mater. 2013, 23, 1466–1471. (23) Siebentritt, S. Why Are Kesterite Solar Cells Not 20% Efficient? Thin Solid Films 2013, 535, 1–4. (24) Smith, W. A.; Sharp, I. D.; Strandwitz, N. C.; Bisquert, J. Interfacial Band-Edge Energetics


33


Page 34 of 41

for Solar Fuels Production. Energy Environ. Sci. 2015, 8, 2851–2862. (25) Neuschitzer, M.; Marquez, J.; Giraldo, S.; Dimitrievska, M.; Placidi, M.; Forbes, I.; Izquierdo-Roca, V.; Pérez-Rodriguez, A.; Saucedo, E. VOC Boosting and Grain Growth Enhancing Ge-Doping Strategy for Cu2ZnSnSe4 Photovoltaic Absorbers. J. Phys. Chem. C 2016, 120, 9661–9670. (26) Guo, Q.; Ford, G. M.; Yang, W.-C.; Hages, C. J.; Hillhouse, H. W.; Agrawal, R. Enhancing the Performance of CZTSSe Solar Cells with Ge Alloying. Sol. Energy Mater. Sol. Cells 2012, 105, 132–136. (27) Bag, S.; Gunawan, O.; Gokmen, T.; Zhu, Y.; Mitzi, D. B. Hydrazine-Processed GeSubstituted CZTSe Solar Cells. Chem. Mater. 2012, 24, 4588–4593. (28) Morihama, M.; Gao, F.; Maeda, T.; Wada, T. Crystallographic and Optical Properties of Cu2Zn(Sn1− xGex)Se4 Solid Solution. Jpn. J. Appl. Phys. 2014, 53, 04ER09. (29) Kim, S.; Kim, K. M.; Tampo, H.; Shibata, H.; Matsubara, K.; Niki, S. Ge-Incorporated Cu2ZnSnSe4 Thin-Film Solar Cells with Efficiency Greater than 10%. Sol. Energy Mater. Sol. Cells 2016, 144, 488–492. (30) Khadka, D. B.; Kim, J. Band Gap Engineering of Alloyed Cu2ZnGexSn1–xQ4 (Q = S,Se) Films for Solar Cell. J. Phys. Chem. C 2015, 119, 1706–1713. (31) Schleich, D. M.; Wold, A. Optical and Electrical Properties of Quarternary Chalcogenides. Mater. Res. Bull. 1977, 12, 111–114. (32) Shu, Q.; Yang, J.-H.; Chen, S.; Huang, B.; Xiang, H.; Gong, X.-G.; Wei, S.-H. Cu2Zn(Sn,Ge)Se4 and Cu2Zn(Sn,Si)Se4 Alloys as Photovoltaic Materials: Structural and


34

Page 35 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Electronic Properties. Phys. Rev. B 2013, 87, 115208. (33) Jackson, P.; Wuerz, R.; Hariskos, D.; Lotter, E.; Witte, W.; Powalla, M. Effects of Heavy Alkali Elements in Cu(In,Ga)Se2 Solar Cells with Efficiencies up to 22.6%. Phys. Status Solidi RRL 2016, 10, 583–586. (34) Minemoto, T.; Matsui, T.; Takakura, H.; Hamakawa, Y.; Negami, T.; Hashimoto, Y.; Uenoyama, T.; Kitagawa, M. Theoretical Analysis of the Effect of Conduction Band Offset of Window/CIS Layers on Performance of CIS Solar Cells Using Device Simulation. Sol. Energy Mater. Sol. Cells 2001, 67, 83–88. (35) Gloeckler, M.; Sites, J. R. Efficiency Limitations for Wide-Band-Gap Chalcopyrite Solar Cells. Thin Solid Films 2005, 480–481, 241–245. (36) Chen, S.; Walsh, A.; Yang, J.-H.; Gong, X. G.; Sun, L.; Yang, P.-X.; Chu, J.-H.; Wei, S.-H. Compositional Dependence of Structural and Electronic Properties of Cu2ZnSn(S,Se)4. Phys. Rev. B 2011, 83, 125201. (37) Paier, J.; Asahi, R.; Nagoya, A.; Kresse, G. Cu2ZnSnS4 as a Potential Photovoltaic Material: A Hybrid Hartree-Fock Density Functional Theory Study. Phys. Rev. B 2009, 79, 115126. (38) Chen, S.; Gong, X. G.; Walsh, A.; Wei, S.-H. Electronic Structure and Stability of Quaternary Chalcogenide Semiconductors Derived from Cation Cross-Substitution of II-VI and I-III-VI2 Compounds. Phys. Rev. B 2009, 79, 165211. (39) Nakamura, S.; Maeda, T.; Wada, T. Electronic Structure of Stannite-Type Cu2ZnSnSe4 by First Principles Calculations. Phys. Status Solidi 2009, 6, 1261–1265. (40) Hartnauer, S.; Körbel, S.; Marques, M. A. L.; Botti, S.; Pistor, P.; Scheer, R. Research Update: Stable Single-Phase Zn-Rich Cu2ZnSnSe4 through In Doping. APL Mater. 2016, 4,


35


Page 36 of 41

070701. (41) Yuan, Z.-K.; Chen, S.; Xiang, H.; Gong, X.-G.; Walsh, A.; Park, J.-S.; Repins, I.; Wei, S.H. Engineering Solar Cell Absorbers by Exploring the Band Alignment and Defect Disparity: The Case of Cu- and Ag-Based Kesterite Compounds. Adv. Funct. Mater. 2015, 25, 6733–6743. (42) Haight, R.; Barkhouse, A.; Gunawan, O.; Shin, B.; Copel, M.; Hopstaken, M.; Mitzi, D. B. Band Alignment at the Cu2ZnSn(SxSe1−x)4/CdS Interface. Appl. Phys. Lett. 2011, 98, 253502. (43) Nagai, T.; Kim, S.; Tampo, H.; Kim, K. M.; Shibata, H.; Takaki, S.; Kawasaki, K.; Kawamura, S.; Shimamura, T.; Matsubara, K.; Niki, S.; Terada, N. Band Alignment of CdS/Cu2ZnSnSe4 Heterointerface and Solar Cell Performances. MRS Adv. 2017, 2, 3157–3162. (44) Udaka, Y.; Takaki, S.; Isowaki, K.; Nagai, T.; Kim, K. M.; Kim, S.; Tampo, H.; Shibata, H.; Matsubara, K.; Niki, S.; Sakai, N.; Kata, T.; Sugimoto, H.; Terada, N. Electronic Structure of Cu2ZnSn(SxSe1−x)4 Surface and CdS/Cu2ZnSn(SxSe1−x)4 Interface. Phys. Status Solidi C 2017, 14, 1600178. (45) Yan, C.; Liu, F.; Song, N.; Ng, B. K.; Stride, J. A.; Tadich, A.; Hao, X. Band Alignments of Different Buffer Layers (CdS, Zn(O,S), and In2S3) on Cu2ZnSnS4. Appl. Phys. Lett. 2014, 104, 173901. (46) Neuschitzer, M.; Lienau, K.; Guc, M.; Barrio, L. C.; Haass, S.; Prieto, J. M.; Sanchez, Y.; Espindola-Rodriguez, M.; Romanyuk, Y.; Perez-Rodriguez, A.; Izquierdo-Roca, V.; Saucedo, E. Towards High Performance Cd-Free CZTSe Solar Cells with a ZnS(O,OH) Buffer Layer: The Influence of Thiourea Concentration on Chemical Bath Deposition. J. Phys. D. Appl. Phys. 2016, 49, 125602. (47) Santoni, A.; Biccari, F.; Malerba, C.; Valentini, M.; Chierchia, R.; Mittiga, A. Valence Band


36

Page 37 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Offset at the CdS/Cu2ZnSnS4 Interface Probed by X-Ray Photoelectron Spectroscopy. J. Phys. D. Appl. Phys. 2013, 46, 175101. (48) Nagai, T.; Udaka, Y.; Takaki, S.; Isowaki, K.; Kawamura, S.; Kawasaki, K.; Tampo, H.; Kim, K. M.; Kim, S.; Shibata, H.; Matsubara, K.; Niki, S.; Terada, N. Electronic Structures of Cu2ZnSnSe4 Surface and CdS/Cu2ZnSnSe4 Heterointerface. Jpn. J. Appl. Phys. 2017, 56, 065701. (49) Bär, M.; Schubert, B.-A.; Marsen, B.; Wilks, R. G.; Pookpanratana, S.; Blum, M.; Krause, S.; Unold, T.; Yang, W.; Weinhardt, L.; Heske, C.; Schock, H.-W. Cliff-like Conduction Band Offset and KCN-Induced Recombination Barrier Enhancement at the CdS/Cu2ZnSnS4 Thin-Film Solar Cell Heterojunction. Appl. Phys. Lett. 2011, 99, 222105. (50) Terada, N.; Yoshimoto, S.; Chochi, K.; Fukuyama, T.; Mitsunaga, M.; Tampo, H.; Shibata, H.; Matsubara, K.; Niki, S.; Sakai, N.; Katou, T.; Sugimoto, H. Characterization of Electronic Structure of Cu2ZnSn(SxSe1−x)4 Absorber Layer and CdS/Cu2ZnSn(SxSe1−x)4 Interfaces by inSitu Photoemission and Inverse Photoemission Spectroscopies. Thin Solid Films 2015, 582, 166– 170. (51) Crovetto, A.; Hansen, O. What is the Band Alignment of Cu2ZnSn(S,Se)4 Solar Cells? Sol. Energy Mater. Sol. Cells 2017, 169, 177–179. (52) Nishihara, H.; Maeda, T.; Shigemi, A.; Wada, T. First-Principles Study of Defect Formation in the Photovoltaic Semiconductor Cu2GeSe4. Jpn. J. Appl. Phys. 2018, 57, 02CE06. (53) Nakamura, S.; Maeda, T.; Wada, T. Phase Stability and Electronic Structure of In-Free Photovoltaic Materials: Cu2ZnSiSe4, Cu2ZnGeSe4, and Cu2ZnSnSe4. Jpn. J. Appl. Phys. 2010,


37


Page 38 of 41

49, 121203. (54) Kumar, S.; Sharma, D. K.; Joshi, B.; Auluck, S. Theoretical Insights into Kesterite and Stannite Phases of Cu2(Sn1–XGeX)ZnSe4 Based Alloys: A Prospective Photovoltaic Material. AIP Adv. 2016, 6, 125303. (55) Tsuji, K.; Maeda, T.; Wada, T. Optical Properties and Electronic Structures of Cu2ZnSnS4, Cu 2 ZnGeS4, and Cu2Zn(Ge,Sn)S4 and Cu2Zn(Ge,Sn)Se4 Solid Solutions. Jpn. J. Appl. Phys. 2018, 57, 08RC21. (56) Collord, A. D.; Hillhouse, H. W. Germanium Alloyed Kesterite Solar Cells with Low Voltage Deficits. Chem. Mater. 2016, 28, 2067–2073. (57) Erdman, P. W.; Zipf, E. C. Low‐voltage, High‐current Electron Gun. Rev. Sci. Instrum. 1982, 53, 225–227. (58) Kraut, E. A.; Grant, R. W.; Waldrop, J. R.; Kowalczyk, S. P. Precise Determination of the Valence-Band Edge in X-Ray Photoemission Spectra: Application to Measurement of Semiconductor Interface Potentials. Phys. Rev. Lett. 1980, 44, 1620–1623. (59) Cozza, D.; Ruiz, C. M.; Duché, D.; Simon, J. J.; Escoubas, L. Modeling the Back Contact of Cu2ZnSnSe4 Solar Cells. IEEE J. Photovoltaics 2016, 6, 1292–1297. (60) Tampo, H.; Kim, K. M.; Kim, S.; Shibata, H.; Niki, S. Improvement of Minority Carrier Lifetime and Conversion Efficiency by Na Incorporation in Cu2ZnSnSe4 Solar Cells. J. Appl. Phys. 2017, 122, 023106. (61) Mitzi, D. B.; Gunawan, O.; Todorov, T. K.; Wang, K.; Guha, S. The Path Towards a HighPerformance Solution-Processed Kesterite Solar Cell. Sol. Energy Mater. Sol. Cells 2011, 95,


38

Page 39 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1421–1436. (62) Richter, M.; Hammer, M.; Sonnet, T.; Parisi, J. Bandgap Extraction from Quantum Efficiency Spectra of Cu(In,Ga)Se2 Solar Cells with Varied Grading Profile and Diffusion Length. Thin Solid Films 2017, 633, 213–217. (63) Persson, C. Electronic and Optical Properties of Cu2ZnSnS4 and Cu2ZnSnSe4. J. Appl. Phys. 2010, 107, 053710. (64) Ley, L.; Pollak, R. A.; McFeely, F. R.; Kowalczyk, S. P.; Shirley, D. A. Total Valence-Band Densities of States of III-V and II-VI Compounds from X-Ray Photoemission Spectroscopy. Phys. Rev. B 1974, 9, 600–621. (65) Cadmium Sulfide (CdS) Band Structure, Energy Gap, Cubic Modification; Springer Materials – Landolt-Börnstein Database, Vol.III/17B-22A-41B, edited by O. Madelung, U. Rossler, and M. Schulz (Springer-New Yorlk, 1999). (66) Yodo, T; Tanaka, S. Growth of CdS Atmospheric Pressure Metalorganic Vapor-Phase Epitaxy at Low Temperature. J. Appl. Phys. 1992, 72, 2781. (67) Nagai, T.; Kanemitsu, Y.; Ando, M.; Kushida, T.; Nakamura, S.; Yamada, Y.; Taguchi, T. Optical Properties of Cubic CdS. Phys. Status Solidi B 2002, 229, 611–614. (68) Kanemitsu, Y.; Nagai, T.; Kushida, T.; Nakamura, S.; Yamada, Y.; Taguchi, T. Free Excitons in Cubic CdS Films. Appl. Phys. Lett. 2002, 80, 267–269. (69) Weinhardt, L.; Fuchs, O.; Groß. D.; Storch, G.; Umbach, E.; Dhere, G. N.; Kadam, A. A.; Kulkarmi, S. S.; Heske, C. Band Alignment at the CdS/Cu(In,Ga)S2 Interface in Thin-Film Solar Cells. Appl. Phys. Lett. 2005, 86, 062109.


39


Page 40 of 41

(70) Dudchak, I. V.; Piskach, L. V. Phase Equilibria in the Cu2SnSe3–SnSe2–ZnSe System. J. Alloys Compd. 2003, 351, 145–150. (71) Nakane, A.; Tampo, H.; Tamakoshi, M.; Fujimoto, S.; Kim, K. M.; Kim, S.; Shibata, H.; Niki, S.; Fujiwara, H. Quantitative Determination of Optical and Recombination Losses in ThinFilm Photovoltaic Devices Based on External Quantum Efficiency Analysis. J. Appl. Phys. 2016, 120, 064505. (72) Gunder, R.; Márquez-Prieto, J. A.; Gurieva, G.; Unold, T.; Schorr, S. Structural Characterization of Off-Stoichiometric Kesterite-Type Cu2ZnGeSe4 Compound Semiconductors: From Cation Distribution to Intrinsic Point Defect Density. CrystEngComm 2018, 20, 1491– 1498. (73) Bekaert, J.; Saniz, R.; Partoens, B.; Lamoen, D. Native Point Defects in CuIn1-xGaxSe2: Hybrid Density Functional Calculations Predict the Origin of p- and n-type Conductivity. Phys. Chem. Chem. Phys. 2014, 16, 22299–22308. (74) Turcu, M.; Kötschau, M. I.; Rau, U. Composition Dependence of Defect Energies and Band Alignments in the Cu(In1-xGax)(Se1-ySy)2 Alloy System. J. Appl. Phys. 2002, 91, 1391–1399.


40

Page 41 of 41 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC GRAPHIC


41

Cu2Zn(Sn1-xGex)Se4 Heterointerface and

Recommend Documents